Predictions & perceptions: A social-ecological analysis of human-carnivore conflict in Botswana

Human-carnivore conflict in agrarian landscapes poses a significant threat to both large carnivore populations and human livelihoods. Conservation strategies aimed at mitigating conflict often rely primarily on ecological, landscape-scale factors. However, assessing the influence of microhabitat, as well as local variation in human perceptions, can offer valuable insights into the spatial patterns of human-carnivore conflict. The objectives of this study were to identify landscape predictors of livestock depredation risk, characterize microhabitat vegetation patterns at kill sites, and assess the spatial alignment between ecologically-predicted and human-perceived depredation risk. Through use of predation risk modeling, microhabitat pattern analysis, and participatory mapping exercises, we provide insight into the spatial patterns of livestock depredation risk from a multi-carnivore guild in an agrarian landscape located in the Makgadikgadi region of Botswana. We identified 170 depredation events and conducted 63 participatory mapping exercises across community lands situated between two wildlife protected areas. Results indicated anthropogenic landscape variables as the primary drivers of general depredation risk. Notably, distinct spatial patterns emerged between African lion (Panthera leo) and African wild dog (Lycaon pictus) depredation, underscoring the need for species-specific conservation strategies. Microhabitat analysis indicated a preference for lower vegetation cover within the multi-predator guild, suggesting livestock microhabitat risk-mitigation behaviors outweigh predator hunting strategies. While human perceptions generally aligned with ecologically predicted risk, identified divergences emphasize the importance of tailored conflict mitigation strategies to accommodate varying influences. Overall, the findings advocate for species-specific depredation risk analyses and the inclusion of human perceptions for a comprehensive spatial understanding of livestock depredation risk.